linux-kernel - [PATCH V2] DT: rockchip: Fix stability issues with the 'performance' governor on rock960

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 17 Oct 2018 11:18:30 +0200
From:   Daniel Lezcano <daniel.lezcano@...aro.org>
To:     heiko@...ech.de
Cc:     linux-kernel@...r.kernel.org,
        Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>,
        Rob Herring <robh+dt@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        Vicente Bergas <vicencb@...il.com>,
        Ezequiel Garcia <ezequiel@...labora.com>,
        Shawn Lin <shawn.lin@...k-chips.com>,
        Enric Balletbo i Serra <enric.balletbo@...labora.com>,
        devicetree@...r.kernel.org (open list:OPEN FIRMWARE AND FLATTENED
        DEVICE TREE BINDINGS),
        linux-arm-kernel@...ts.infradead.org (moderated list:ARM/Rockchip SoC
        support),
        linux-rockchip@...ts.infradead.org (open list:ARM/Rockchip SoC support)
Subject: [PATCH V2] DT: rockchip: Fix stability issues with the 'performance' governor on rock960

When the performance governor is set as default, the rock960 hangs
around one minute after booting, whatever the activity is (idle, key
pressed, loaded, ...).

Based on the commit log found at https://patchwork.kernel.org/patch/10092377/

"vdd_log has no consumer and therefore will not be set to a specific
voltage. Still the PWM output pin gets configured and thence the vdd_log
output voltage will changed from it's default. Depending on the idle
state of the PWM this will slightly over or undervoltage the logic supply
of the RK3399 and cause instability with GbE (undervoltage) and PCIe
(overvoltage). Since the default value set by a voltage divider is the
correct supply voltage and we don't need to change it during runtime we
remove the rail from the devicetree completely so the PWM pin will not
be configured."

After removing the vdd-log from the rock960's specific DT, the board
does no longer hang and shows a stable behavior.

Apply the same change for the rock960 by removing the vdd-log from the
DT.

Fixes: 874846f1fccd ("arm64: dts: rockchip: add 96boards RK3399 Ficus board")
Tested-by: Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>
Signed-off-by: Daniel Lezcano <daniel.lezcano@...aro.org>
---
Changelog
  - V2
     * Added Tested and Fixes tags
---
 arch/arm64/boot/dts/rockchip/rk3399-rock960.dtsi | 12 ------------
 1 file changed, 12 deletions(-)

diff --git a/arch/arm64/boot/dts/rockchip/rk3399-rock960.dtsi b/arch/arm64/boot/dts/rockchip/rk3399-rock960.dtsi
index 6c8c4ab..56abbb0 100644
--- a/arch/arm64/boot/dts/rockchip/rk3399-rock960.dtsi
+++ b/arch/arm64/boot/dts/rockchip/rk3399-rock960.dtsi
@@ -57,18 +57,6 @@
 		regulator-always-on;
 		vin-supply = <&vcc_sys>;
 	};
-
-	vdd_log: vdd-log {
-		compatible = "pwm-regulator";
-		pwms = <&pwm2 0 25000 0>;
-		regulator-name = "vdd_log";
-		regulator-min-microvolt = <800000>;
-		regulator-max-microvolt = <1400000>;
-		regulator-always-on;
-		regulator-boot-on;
-		vin-supply = <&vcc_sys>;
-	};
-
 };

 &cpu_l0 {
-- 
2.7.4